How to Form A Survey Effectively
This is a question that frequently enters the mind of non-profit program managers, healthcare professionals, product managers, school principals, the list goes on. The key word to this title, however, is how does one do this effectively? Sure, asking questions is easy, but how do you know they’re even the right questions?
Top Considerations: Where to Begin
1. What’s the purpose of the survey? This is the most important question. What questions are you trying to answer? Are you trying to eventually publish this somewhere? If so, the design needs to be particularly rigorous.
2. How do you measure these answers? For this question, ideally you use a pre-existing questionnaire. Questionnaires need to be validated and reliable for them to be relevant. This ensures your format didn’t unduly influence the individual taking the survey and can be repeated. This is an incredibly important question, which we’ll go into more depth below.
3. Has this been done before? Are you drawing new answers by doing this survey? Can you learn from previous attempts? Are there experts in this area who you could consult?
4. Who are you trying to reach? When we form surveys, we want to keep the individual respondent in mind. Does the survey need to be tailored towards their particular condition, age-range, or other demographic?
5. Does the format you have in mind work for your respondents? It’s critical that the survey you’re designing is going to be accessible to the individuals you’re trying to survey. Is paper and pen in-office the best method? Over the phone? Or is a more technical format that streamlines survey data into one database like a cell phone application most accessible for your respondents?
6. Data cleaning, analysis, and dissemination — Who will look at the data? Having data isn’t helpful unless someone is looking at it, drawing conclusions, and synthesizing it. What are the questions that will be helpful to your audience, and who is the designated team to help aggregate conclusions about the data. Can they organize the data, clean it appropriately, recognize outliers, and visualize conclusions succinctly?
Survey Factors: Considerations in Writing
1. Item-order Consider what order the questions are going to appear, and whether they may affect responses on the following ones. This is particularly key for valenced questions that may represent evaluative judgment on self for an individual, for example: life satisfaction, how often individuals feel a certain way, etc.
2. Types of items Open-ended items allow minimal influence of the research on the patient, but trade-off with the difficulty of coding and transcribing. Individuals are more likely to skip these questions because they take longer. Questions where you don’t want to influence the participant or force them into a box, should be open-ended. Do you want to use a 1–5 Likert scale that allows you to range from ‘Very Poor’ to ‘Excellent’? Do you need to judge measurable differences like a 1–10 numerical scale? Would a visual analog scale with happy to unhappy faces work best? On the other hand, yes or ‘I agree’ vs ‘no’ or ‘I disagree’ are straightforward and don’t allow for ambiguity. Participants are forced to respond in a valenced capacity.
3. Writing Effective Items: BRUSO: Authors Price, Jhangiani & Chiang use this acronym that stands for “brief,” “relevant,” “unambiguous,” “specific,” and “objective.” Additionally, are your items:
- To the point without extra language
- Relevant to the research question
- Use scales that are familiar to individuals when possible.
Lastly, be sure to introduce your survey in some capacity. This both encourages participation and its explains the purpose of the survey, and allows for a fluid touchpoint to gain consent for their answers.
Price, Paul C., Jhangiani, Rajiv S. & Chiang, I-Chant A. “Constructing survey questionnaires.” Research methods in psychology (2015). https://opentextbc.ca/researchmethods/chapter/constructing-survey-questionnaires/